Premature rupturing of fetal membranes (the membrane defining the amniotic sac) in pregnant women occurs in about 5% to 14% of all pregnancies and is the cause of about 10% of all perinatal deaths. When a fetal membrane rupture occurs toward the end of a pregnancy (38-40 weeks of pregnancy), delivery of the fetus should be effected as soon as possible in order to minimize the associated risks to the fetus and the mother. A simple, rapid and effective method for identifying fetal membrane ruptures is therefore needed so that the occurrence of a fetal membrane rupture toward the end of a pregnancy can be effectively monitored in order to limit the amount of time between the occurrence of the fetal membrane rupture and delivery.
Premature rupturing of fetal membranes is not a problem that is limited to the last few weeks of pregnancy. More than 30% of premature fetal membrane ruptures occur before 37 weeks of pregnancy. Fetal membrane ruptures before 37 weeks of pregnancy have been associated with significant increases in the risk of an intrauterine infection, Keirse M. J. N. C., et al., "Prelabor Rupture of the Membrane Preterm," in Efficient Care in Pregnancy and Child-birth, 1989, Vol. 1, Oxford, New York, Toronto. Edited by J. Chalmers, M. Enkin, and M. Keirse. Intrauterine penetration of such infections increase both maternal and perinatal mortality. The risk of intrauterine penetration of the intrauterine infection increases significantly as time passes between the rupture of the membrane and delivery. Early diagnosis of fetal membrane ruptures that occur before 37 weeks of pregnancy is therefore extremely important. Positive diagnosis of fetal membrane ruptures before 34 weeks of pregnancy is also important since it permits the timely monitoring and treatment of pregnant women to prevent intra-amnion infection and to stimulate fetal lung development.
A variety of methods have been developed for detecting fetal membrane ruptures. One method, called the crystallization test, detects amniotic fluid in vaginal secretions based on the observation of arborization, a tree-branch-like structure which forms when amniotic fluid dries on a slide. The crystallization test is described in M. L. Friedman and T. W. McElvin, American Journal of Obstetrical Gynecology, 1969, Vol. 104, pp. 544-550.
The crystallization test relies on the visual image produced on a slide by a vaginal secretion containing amniotic fluid being sufficiently visually distinct from normal vaginal secretions so as to enable the detection of amniotic fluid in a vaginal secretion. Because the crystallization test relies on the user to identify the presence of arborization in a test sample, the crystallization test is highly susceptible to user error. As a result, the accuracy of the crystallization test can be significantly operator dependent.
Many factors can cause a vaginal secretion sample to appear as though it contains amniotic fluid when it does not, and visa versa. For example, results obtained from the crystallization test can be erroneous if a long time has passed since the rupture has taken place. Vaginal infections can also influence the results of the crystallization test. It has been found that the crystallization test produces false results in as many as 20% of the cases, most commonly due to the pattern produced by a vaginal secretion being misidentified as containing amniotic fluid or not containing amniotic fluid.
A method has also been proposed for detecting fetal membrane ruptures based on the pH of a vaginal secretion where a detectable agent, such as a dye, is used to stain the amniotic fluid in vaginal secretion samples. M. L. Friedman and T. W. McElvin, American Journal of Obstetrical Gynecology, 1969, Vol. 104, pp. 544-550. Detection of fetal membrane ruptures based on the pH of vaginal secretions is inaccurate due to the susceptibility of the test to user error, as well as the variability of the pH of vaginal secretion samples person to person due to factors independent of the presence or absence of amniotic fluid in a vaginal secretion.
Methods have also been proposed for detecting fetal membrane ruptures based on an immunochemical analysis of the proteins contained in amniotic fluid. Immunochemistry is a branch of science that deals with the chemical changes and phenomena of immunity, specifically, the chemistry of antigens, antibodies, and their reactions. The proposed methods utilize the following four protein compounds for diagnosing fetal membrane ruptures: alpha-fetoprotrin, prolactin, fibronectin, and insulin-like growth-factor-binding protein 1 (IGFBP-1). B. L. Rochelson, et al., "Rapid Assay-Possible Application in the Diagnosis of Premature Rupture of the Membranes," in Obstetr. Gynecol., 1983, Vol. 62, pp. 414-418; Koninckx, et al., "Prolactin Concentration in Vaginal Fluid: A New Method for Diagnosing Ruptured Membranes," Br. J. Obstetr. Gynecol. 1981, Vol. 88, pp. 607-610; P. Hellemans, et al., "Preliminary Results with the Use of Rom-Check Immunoassay in the Early Detection of Rupture of the Amniotic Membranes," Eur. J. Obstetr. Gynecol. Reprod. Biol., 1992, Vol. 43, pp. 173-179; Rutanen, E. M., et al., "Measurement of Insulin-Like Growth-Factor Binding Protein-1 in Cervical/Vaginal Secretions: Comparison with the ROM-Check Membrane Immunoassay in the Diagnosis of Ruptured Fetal Membranes," Clin. Chim. Acta, 1993, Vol. 214, pp. 73-81).
Among the above immunochemical methods, those which are based on the detection of alpha-fetoprotein (AFP) and prolactin (PRL) in vaginal secretions are unreliable because the corresponding blood/amniotic fluid ratio of the above proteins varies considerably, i.e., between 1 and 10. In some cases, higher concentrations of proteins have been found in serum than in amniotic fluid. In addition, the concentration of amniotic proteins, such as alpha-fetoprotein, prolactin, and fibronectin, in amniotic fluid also varies during the course of a pregnancy. For example, AFP and PRL are present in amniotic fluid in high concentrations only during the second trimester (i.e., the second three-month period) of the pregnancy. As the pregnancy advances, the amniotic/serum protein concentration ratios for both these proteins decrease and is only about 3 to 4 at term.
Detection of fetal membrane ruptures based on the presence of fetal fibronectin in vaginal secretions has also been found to be unsatisfactory. P. Hellemans, et al., "Preliminary Results with the Use of the ROM-Check Immunoassay in the Early Detection of Rupture of the Amniotic Membranes," Eur. J. Obstetr. Gynecol. Reprod. Biol., 1992, Vol. 43, pp. 173-179; C. Lockwood, et al., "Fetal Fibronectin in Cervical and Vaginal Secretions as a Predictor of Preterm Delivery," in The New England Journal of Medicine, 1991, Vol. 325, pp. 669-674. The ROM-Check immunoassay taught by Hellemans, et al. is an immunochemical method based on the detection of fetal fibronectin.
Rutanen, et al. teaches an immunochemical method based on the detection of insulin-like growth-factor binding protein-1 (IGFBP-1). Rutanen, E. M., et al., "Measurement of Insulin-Like Growth-Factor Binding Protein-1 in Cervical/Vaginal Secretions: Comparison with the ROM-Check Membrane Immunoassay in the Diagnosis of Ruptured Fetal Membranes," Clin. Chim. Acta, 1993, Vol. 214, pp. 73-81). In this study, Rutanen, etal. teaches a ROM-Check membrane immunoassay with a false positive rate of 20% and a false negative rate of 9%.
The aforementioned methods for detecting fetal membrane ruptures based on the detection of alpha-fetoprotein, prolactin, fibronectin, and insulin-like growth-factor-binding protein 1 are not highly accurate, due, at least in part, to the many variable factors regarding the concentration of these proteins in amniotic fluid, and the relative concentration of these proteins in amniotic fluid to serum. The natural occurrence of these proteins in serum, albeit at a significantly lower concentration, creates a background noise level which can lead to the false identification of a fetal membrane rupture or the absence thereof. The background noise level due to the natural occurrence of these proteins in serum is particularly significant when the amount of amniotic fluid in the vaginal secretion sample is small. As a result, these methods are inaccurate and unreliable for the detection of fetal membrane ruptures.
At least one method has been developed for detecting fetal membranes ruptures based on the presence of IGFBP-1 in vaginal secretion samples at higher levels than is normally found in serum. International Patent Application WO 92/12426 to Eeva-Marja Rutanen, 1992. Two monoclonal antibodies (MAb 6303 and MAb 6305) which are capable of binding to IGFBP-1 are used in this method to determine the total amount of IGFBP-1 in a vaginal secretion sample. As will be discussed herein, antibodies MAb 6303 and MAb 6305 recognize functionally different antigenic determinants on the surface of the IGFBP-1 molecule than the antibodies of the present invention. It is the different antigenic determinants of the antibodies of the present invention which crucially determines their medical interest and their utility in the methods and devices of the present invention.
The method described in WO 92/12426 uses a method known as a two-site immunoradiometric assay which is described by F. Pekonen, et al., in Journal of Immunoassay, 1989, Vol. 10, pp. 325-337. More specifically, the method involves placing a vaginal secretion sample into a sample-holding plate containing one of the two monoclonal antibodies (MAb 6303 and MAb 6305). IGFBP-1 molecules contained in the vaginal secretion sample are attached to a first of the two antibodies which is present in the holding plate. A specially-labeled second antibody is then introduced and is connected to another site of the same IGFBP-1 molecule. Those labeled second antibodies which bind to the IGFBP-1 molecule are immobilized on the holding plate and are subsequently measured by methods known in the art, for example, by means of a radioactive counter.
Using the method described in WO 92/12426, one can quantitatively determine the amount of IGFBP-1 in a vaginal secretion sample. This method partially overcomes the disadvantages associated with the above described immunochemical methods by selecting a protein for analysis in which the blood/amniotic fluid ratio varies over the second and third trimesters within a narrower range than the immunochemical methods described above.
One serious drawback of the diagnostic method proposed by Rutanen et al. is that IGFBP-1 can be present in the blood in relatively large concentrations. For example, the concentration of all forms of IGFBP-1 in the sera of pregnant women ranges from 58 to 600 ng/mL (median 220 ng/mL). As a result, even small admixtures of serum can cause an increase in the level of IGFBP-1 detected in the sample to be significantly higher than the level of sensitivity (about 0.5 ng/mL) of the Rutanen method. For example, IGFBP-1 can be present in a vaginal secretion sample in a concentration range from 0.5 to 90 ng/mL in women with an intact fetal membrane. In addition, Rutanen assumes that even with the fetal membrane being intact, a trace amount of IGFBP-1 is still leaking into the vagina.
In order to reduce the frequency of false positive results in the Rutanen method based on the presence of a low concentration of IGFBP-1 in a vaginal secretion not due to a fetal membrane rupture, (for example, due to blood or IGFBP-1 leaking into the vagina from an intact membrane), the Rutanen method requires a relatively high concentration of IGFBP-1 to be detected (100 ng/mL) before being considered to indicate the presence of a fetal membrane rupture. For example, Rutanen treats the occurrence of a lower concentration of IGFBP-1 in a sample than the highest known concentration of IGFBP-1 in maternal serum as a determination that a fetal membrane rupture has not occurred. This approach, however, could cause the method described in WO 92/12426 to produce a high level of false negative results. For example, when the rupture is small or in its initial stage, the small increase in IGFBP-1 in the vaginal secretion relative to serum may not be sufficient to be recognized by the Rutanen method as a positive result. This severely limits the Rutanen method as a reliable early detection method.
All of the methods described above can be used to detect detect the presence of a fetal membrane rupture based on the detection of a biomolecule in a vaginal secretion which occurs in amniotic fluid at a higher level than in serum. A common disadvantage to all of these methods is the fact that the biomolecule being detected can also be present in appreciable amounts in serum. As a result, an unacceptably high level of false positive results can occur when the concentration of the biomolecule in the patient's serum is unusually high and an unacceptably high level of false negative results can occur when only a very small amount of amniotic fluid is present in the vaginal secretion.
A need therefore exists for a method for detecting fetal membrane ruptures based on the presence of a biomolecule in a vaginal secretion which occurs in amniotic fluid at a significantly higher level than in serum. The biomolecule should also occur in a lower concentration in serum than the biomolecules which have been previously used to detect fetal membrane ruptures.
A need also exists for a method which can detect fetal membrane ruptures based on the presence of amniotic fluid in a vaginal secretion with a high level of accuracy. In order to minimize the number of false negative results, it is important that the method be able to detect small concentrations of amniotic fluid in a vaginal secretion.
The above described methods are also designed generally for measuring the presence of a biomolecule in a relatively narrow range of its concentration. For example, some methods require dilution of the sample prior to testing. In order to broaden this range, multiple attempts to match the concentrations of the sample to the specific range of protein concentrations must be made. A need exists for a method which can be used over a wide range of biomolecule concentrations.
A further common disadvantage of the above-described methods is the significant amount of time required to perform these methods, as well as the amount of laboratory equipment and skill required to perform these methods. A need exists for a method which can be rapidly performed without appreciable laboratory equipment. A need also exists for a method whose accuracy is not operator dependent and can be performed by untrained personnel, such as the patient.
A variety of test devices which are based on visual color detection of various antigens are known. European Patent Application 421,294 A2 to E. Osikowic, 1991. However, a simple test kit suitable for rapidly detecting the presence of amniotic fluid in a vaginal secretion is not known. Although Rutanen specifies in WO 92/12426 that a special kit has been developed which may be used to carry out the method, no description of any test kit, except for the reagents used in the method for measuring the concentration of IGFBP-1 is provided. There are no drawings or any other physical description of the kit as a device. A need therefore exists for a test device and kit for detecting the presence of amniotic fluid in vaginal secretions.